Eric Watkins argues that a grasp of Leibnizian and anti-Leibnizian thought in eighteenth-century Germany helps one to see how Kant (in his critical period) argued for causal principles that have both metaphysical and epistemological elements. According to this interpretation, Kant's model of causality does not consist of events, but rather of substances endowed with causal powers that are exercised according to their natures and circumstances.
This book is not for the beginner in philosophy! But it is greatly rewarding for those who already have a grasp of Kant's body of work. Watkins identifies a model of causality that begins in the debates over causality between Leibnizians and anti-Leibnizians. The former believe in a theory of pre-established harmony, in which individual substances are causally isolated but nevertheless appear coordinated through the intervention of God. Kant positions himself as a defender of physical influx, which is the idea that substances interact with one another. Mutual interaction of substances is required for change, since a substance is immutable in itself. The mere existence of substances, moreover, is not enough to bring about change. Instead, there must be a change in relations between them. Change is therefore accounted for in terms of an activity of substances--i.e. the causality of a cause. This model of causality emerges in the analogies of experience, which Watkins interprets as having both epistemological and metaphysical import insofar as knowledge of temporal relations, and of a single world, require substances in causal relationships with one another. This model of causality involving substances and powers, which is fundamentally distinct from a Humean model involving discrete and indivisible (i.e. instantaneous, since Hume believed in the finite divisibility of time) events, emerges also in Kant's account of laws of nature (they are unchanging insofar as they depend on unchanging grounds of substances) and also free will (which is best understood as a form of agent-causation and altered-law compatibilism). This book might require more than one read to gain a full appreciation of it.
Nothing like this has been written on Kant's model of causality and its metaphysical implications for the Analogies of Experience, the Third Antinomy, and his response to David Hume. Watkins is able to place Kant's work within the philosophical and historical context of his time, a context that differs in important respects from the standard historical account generally given. Though the ideas being expressed are subtle and complex, the work required to break down the material is well rewarded with profound and penetrating insight into Kant's project and the deeper concerns of his work. This is an important read for anyone with more than a casual interest in Kant's philosophy.